Uncertainty quantification in geoelectrical imaging and monitoring

Project description

Over the past decade, geoelectrical imaging has become the leading technology for continuously monitoring the shallow subsurface volumetrically and in real time. This technology plays a crucial role in assisting industrial and governmental stakeholders in addressing some of the most pressing societal challenges that impact on the subsurface, such as unconventional energy sources, carbon sequestration, waste management and groundwater contamination. However, current geoelectrical imaging techniques do not allow appropriate quantification of the uncertainty intrinsic to (1) the subsurface and (2) conventional image reconstruction methods based on deterministic inversion. The absence of uncertainty quantification in the subsurface has profound detrimental effects on evidence-based decision-making, the assessment and management of risks associated with subsurface hazards, the design of cost-effective remediation strategies and the improvement of stakeholder and public acceptance in the context of potentially controversial uses of the subsurface (e.g. unconventional hydrocarbons, CO2 storage). This project will develop Bayesian methodologies for geoelectrical imaging with the ultimate aim of inferring and quantify uncertainty in subsurface properties in the presence of realistic geologies. The project will be focused on applications that include carbon capture and storage, unconventional hydrocarbons and underground gas storage.